Devoutly confessing that you do not know is better than prematurely claiming that you do
“Mark well that upon which you set your heart” said God. “For you shall achieve it”
Other things being equal, what’s there not to like about introducing more choice and competition in areas in which governments fund and/or deliver human services? That’s the starting point of the Productivity Commission’s recently commenced inquiry into competition and choice in human services. These services are so important economically to the government’s bottom line and to Australians’ wellbeing that it gives Australia another opportunity to show global leadership in a difficult area – as we did in the glory days of economic reform from 1983 until the early 2000s when uniquely within the Anglosphere, we pursued ‘neoliberal’ economic efficiencies whilst also strengthening the social safety net.
However, this is difficult territory and ideological debates for and against competition won’t get us far – indeed, they could see us embrace some bold new agenda which somehow fails to identify, let alone tackle, the really knotty problems that pervade the area.
We’re hoping Gary Sturgess’s admonitions at the two-day workshop on the subject run by the Harper Review are still ringing in some ears. Listening to the easy encomiums to competition from one connoisseur of ‘high policy’ after another, Sturgess offered this: “Stop it or you’ll go blind”. Sturgess was unimpressed with the Procrustean way in which the first draft of the Harper Report rehearsed the advantages of competition and choice without any real understanding of the difficulties of delivery – without entering into the messy world of those on the ground.
Too often in Sturgess’s critique, ‘the market’ becomes a deus ex machina somehow absolving the policymaker from their fundamental responsibilities:
“Until very recently, when questions concerning the design and stewardship of the supply side [of the NDIS] were raised, policymakers responded: ‘oh, the market will fix that’. The problem is that there is no market — that is the challenge that policymakers have set. And there is no reason why the market will address supplier failure in a timeframe that will be relevant to a person with a severe disability.” Gary L. Sturgess, 2016. “Reflections on the Public Service Economy”, Annual Lecture of the Business Services Association, London, 7 June.
Government as custodian of a corporate supply chain
Sturgess proposes that the master narrative in the area should not be contracting out, competition, choice or even market design, but rather the management of a corporate supply chain with all of those phenomena being part of the repertoire:
“If Ford mismanages its supply chain, so that its vehicles burst into flames, there is no point blaming their suppliers. It is Ford that will face the cost of recall. It is Ford’s brand that will be trashed. It is Ford’s share price that will suffer. And while Ford may lay off some of the blame on its suppliers, it is Ford that will bear the vast majority of the reputational and financial cost.”
We agree, and think the idea can be taken further.
Cultivating and accessing distributed knowledge and capability
Firstly, Ford was an exemplar of corporate performance around a century ago! In many ways, the critique Friedrich Hayek offered of central planning in government applies to large corporates in the image of the firm Henry Ford built. Just as Sturgess argues that policy valourises high policymaking over the quotidian world of delivery, so Hayek critiques central planners’ inability to access and utilise knowledge (often informal knowledge) distributed throughout the system.
In that regard, the Toyota production system offers additional insights over that of Ford. Developing Henry Ford’s ideas about driving out waste, Toyota understood better than Ford that it needed to do more than simply sit at the centre of the immensely complicated system it bestrode, issuing instructions to those on the production side – its employees and suppliers – and then marketing the result to its customers.
It drew its employees and suppliers into the process of endlessly optimising their own productive performance in myriad ways that couldn’t have been directed from the centre. Where American car companies sent their engineers into suppliers’ facilities to spot efficiency gains their suppliers could make and then captured those gains for themselves by cutting suppliers’ prices, Toyota allowed its suppliers to share the resulting cost reductions. So the innovations kept coming. Toyota also cultivated higher trust relationships with employees, spending an order of magnitude more than its American competitors on training and offering job security and stable salaries (as opposed to performance-based piece rates). It was a similar story with customers. Rather than maximise production and then sell inventory – if necessary with hefty incentives – Toyota built to order and engaged dealerships both in building long-term relationships with customers and also in feeding back market intelligence to the production side.
Competition and transactional market relations were the leitmotif of the American system with the result that the conflicts of interest between the various parts of the system – managers, designers, employees, suppliers, dealers, customers – often dominated, crowding out the common interests they all shared. Though market relations sat in the background conditioning and disciplining relationships, Toyota’s focus was on long-term cooperative relationships between the parts of the whole system.
Information and hierarchy
In all this – just as Hayek emphasised in a very different context – information was critical. It should have been axiomatic since at least the days of Gosplan in Soviet Russia, that information systems imposed from the top are almost invariably gamed. This is the bane of new public management. Just ask the patients of the Bristol Eye Hospital which met its targets from the centre to reduce waiting lists – by delaying reminders for regular eye checkups!
The information system behind Toyota’s production system was, by contrast, an exercise in applied humility. Predicated on the idea that, properly managed and trained, teams of employees have an intrinsic desire to perform well, Toyota built its production information system primarily around the needs and perspectives of those on the line and their accountability to themselves in the first instance. Employees were trained in statistical control and the management of computer numeric control (CNC) machine tools and tasked with using that knowhow to endlessly optimise performance between and across teams in regular ‘quality circle’ meetings. Of course, as well as doing its job on the line, production data was more widely available to management.
Information, transparency and competition
A few decades after Hayek had highlighted the difficulties centrally planned economies have in managing information, scholars such as Arrow, Akerlof, Shiller and Stiglitz did a similar hatchet job on markets. Markets mishandle information for at least two reasons.
- However persuasive Hayek’s points might have been about the power of the price system in signalling the opportunity cost of production and consumption decisions to all players in the market, prices are the only information about which there can be no asymmetry between buyer and seller – for they are agreed between them! But there’s a serpent in this paradise and it’s the same phenomenon that corrupts bureaucratic systems of accountability. Asymmetric information between different players in the market – buyers, sellers and intended beneficiaries of services – corrupts the generation and flow of information about product quality and the various needs and circumstances of the players.
- Information is, in economic jargon, a ‘non-rival’ good. Where a physical good like a car or a cookie is consumed exclusively, non-rival goods like ideas and know-how can be enjoyed by one without interfering with others’ enjoyment of them.
Thus, no practical arrangement for handling information will be perfect in the manner it might be in an economist’s model. We should look to evolve hybrid institutions that capture the best of competition and collaboration where possible, whilst avoiding their worst pathologies. That’s what Toyota did with its supply chain.
In addition to the generous technical support it offered its suppliers, it also sponsored regular supplier wide-open days – both at Toyota and at the supplier factories. Non-cooperation in the knowledge commons was not an option for suppliers wanting to retain Toyota’s patronage. This rapidly normalised the culture of sharing and collaboration both within and between firms within the ‘family’ of suppliers, thus increasing the rate at which successful innovations spread. In the upshot, Toyota’s knowledge and learning commons was instrumental in its often doubling its American competitors’ labour productivity while exceeding their production quality.
A supply chain needs a brain
About two years after the fall of the Berlin Wall, a delegation of Russian bureaucrats visited London and one famously asked Paul Seabright, a leading economist who met with them, “who is in charge of the supply of bread to the population of London?” Counterintuitive though it may be, in a market, no one’s in charge. That’s not true of a corporate supply chain. It will make use of markets in myriad ways, but, like a body, it will have a nervous system providing it continual feedback about the performance of its various parts, and a brain capable of aggregating that information and strategising improvement.
“A ‘brain’ works out what creates outcomes and what doesn’t. A ‘brain’ would know how the system is currently performing and how to improve and grow services when they are working. . . . In many of the sectors in which we work, though the initial system rarely performed as we would like, increased outsourcing has, in fact, reduced expertise as to what works. The risk is that government has become an expert in contract management, and service providers have become experts in ‘contract delivery’. ‘Brains’ have atrophied on both sides. We’d like to see the Commission reverse this trend and enable both to be experts in ‘outcome creation for vulnerable groups’. We work with some of the most progressive government departments and service providers and they often struggle to keep their ‘brains’ intact because of the larger forces operating around them.”
Indeed, in exploring one area of human services that had produced disappointing results within a large department run by senior managers who were dedicated and highly regarded (including by us), we nevertheless found that one operational unit had begun delivering considerably improved performance. But this went undetected amongst senior management and the unit was disbanded with changing departmental priorities.
A brain needs a nervous system: managers must be kept honest
Even if government administrations learn the lessons argued for above, the greatest difficulty remains as it ever was. For profit-seeking firms in a market, quality, cost, profitability and economic value are relatively tangible. Even where they are quite imperfect, markets offer considerable intelligence (transparency) about the economic value of specific goods or services and comparative costs between suppliers, all of which helps keep management to the honest toil of satisfying customers at minimum cost.
Human services are much more difficult – conceptually and practically. Firstly, in the human world, interactions are more complex and uniquely situated within a context than is the case in the market provision of most products. So it’s much harder to know what works, how and why it works, how it affects other parts of the system, the difference between long- and short-term impacts and so on. Secondly, because of their public good and benevolent quality, many human services aren’t funded by paying consumers. This robs the resulting systems of all the transparency and feedback around cost, price and quality that emerges from the organic tension between buyers and sellers in a market.
In lieu of these market signals and disciplines, system managers typically specify performance measures. But designing good performance measures – measures that will have some diagnostic power for those operating the program – will often require intimate knowledge of the workings of the program itself. For instance, an appealing measure of the performance of a job placement program would be the number of job seekers continuing in new jobs after job-placement services. For a child protection program, it might be the number of children removed from struggling families following early intervention. But are these the right measures? Could too rapid job matching destroy value by foreclosing better matches, or by diverting valuable system resources to where they are redundant? And how does one weigh up the relative merits of child removal with poor home care? It is these things that monitoring and evaluation should be shining a light on, enabling the system to better understand and improve its own impact.
In all this, performance measures imposed from the top sound like a mistake waiting to happen. Bureaucracies have a terrible habit of role-playing their expertise, while in reality going through the motions and covering their arses. And this can occur whether service is delivered by lower levels of the hierarchy or by contractors. Yet our experience in TACSI tells us that progress occurs when we draw those we are trying to help into the process, when we’re intentional about the change we’re trying to facilitate and about the process of learning. That means articulating our theory of change and then testing each assumption we’re making about how change occurs. The rigour of this process, against which we’re constantly testing our practice, is ‘scientific’, seeking to test itself against reality at every turn. Competition within functioning markets has the same quality of relentlessly disciplining those within firms against a hard reality from without – from the market.
The real question confronting the PC’s inquiry into Human Services is how one might generalise TACSI’s ideas for program design and management to the system level. We’d propose these building blocks:
- Monitoring and evaluation (knowing what you’re doing) should be at the heart of program delivery – whether delivery is by government or external providers.
- It needs to be designed and administered at the level of those delivering services on the ground.
- However, the ultimate responsibility for the monitoring and evaluation system and the data within it should lie, not with the portfolio minister to whom the delivery agency reports, but with an independent agency that reports directly to Parliament like the Auditor General or the Ombudsman.
Monitoring and evaluation – the ‘nervous system’ of the production chain would thus be:
- Co-designed by those at the coalface of delivery and by experts in monitoring and evaluation;
- Independent of the institutional imperatives of the delivering agency;
- Transparent to all to maximise the scope for the wider system to learn, and to deliver improved bureaucratic and political transparency of performance.
Finally, it is worth noting the way choice is typically framed. It’s certainly worth asking whether allowing firms or NGOs to compete to supply human services might work better than delivery by government hierarchies. But questions about choice should also provoke deeper reflection. Consider how case-managers are chosen for clients in child protection agencies, or teachers for students or nurses for hospital patients. Allowing patients to choose their nurses could substantially disrupt efficient routines, but giving students greater choice in their teachers or clients greater choice of case-workers could prompt all manner of system improvements. This isn’t to say that choice is self-evidently the best criterion by which such matches should be made. These are complex cases and the clients of these services may not be particularly wise in pursuing their enlightened self-interest. However, exploring ways to reframe choice and specifically who chooses what (or whom)within human service agencies seems like a promising and overlooked consideration, whether or not we arrange the system to look like a market by having multiple suppliers compete for business.
In fact, in a TACSI-built peer-to-peer mentoring program for families going through tough times – Family by Family – the family that identifies itself as needing help chooses its mentor family from a selection of profiles as someone might seek a date on an online website. But crucially, we didn’t design this part of the program from some ideological judgement that choice is better, or even that choice would ’empower’ families seeking help – though we did want to empower them. It just seemed natural to do this as we co-designed the program with the families aroundtheir perspectives and needs, not the needs of all the higher status participants in the process – the mentor family, the family coach and other professionals running the program.
As long ago as 1907, John Dewey was ruminating on what might today go under the slogan of ‘citizen centric’ services. Thinking of schools, he wrote:
The change which is coming into our education is the shifting of the center of gravity. It is a change, a revolution, not unlike that introduced by Copernicus when the astronomical center shifted from the earth to the sun. In this case the child becomes the sun about which the appliances of education revolve; he is the center about which they are organized.
In many ways, schools have been transformed just as Dewey hoped – using his own words to drive the change. And yet, compared with an imaginable alternative embodying the spirit and not just the letter of his words, he’d be surprised at how little progress we’ve made at building services around those they’re intended to benefit – in schools, in hospitals, in human services more generally.
We should be wary of a repeat in which, to use Gary Sturgess’s words, in the pursuit of reform we “simply replace one kind of institutional monoculture with another”.
The interview of the essay: